ACTA ASTRONOMICA Vol. 58 (2008) pp. 279–292 New Variable Stars on Digitized Moscow Collection Plates. Field 66 Ophiuchi (Northern Half) D.M. Kolesnikova1 , L.A. Sat2 , K.V. Sokolovsky2,3,4 , S.V. Antipin2,1 and N.N. Samus1,2 1 Institute of Astronomy, Russian Academy of Sciences, 48, Pyatnitskaya Str., Moscow 119017, Russia 2 Sternberg Astronomical Institute, Moscow University, 13, University Ave., Moscow 119992, Russia e-mail: [email protected], [email protected], [email protected] 3 Astro Space Center of Lebedev Physical Institute, Profsoyuznaya 84/32, 117997 Moscow, Russia 4 Currently at: Max Planck Institute for Radio Astronomy, Auf dem Hügel 69, 53121 Bonn, Germany Received June 9, 2008 ABSTRACT We initiated digitization of the Moscow collection of astronomical plates using flatbed scanners. Techniques of photographic photometry of the digital images were applied, enabling an effective search for new variable stars. Our search for new variables among 140000 stars in the 10◦ × 5◦ northern half of the field centered at 66 Oph, photographed with the Sternberg Institute’s 40-cm astrograph in 1976–1995, gave 274 new discoveries, among them: 2 probable Population II Cepheids, 81 eclipsing variables, 5 high-amplitude δ Sct stars (HADSs), 82 RR Lyr stars, 62 red irregular variables and 41 red semiregular stars, 1 slow irregular variable not red in color. Ephemerides were determined for periodic variable stars. We detected about 30 variability suspects for follow-up CCD observations, confirmed 11 stars from the New Catalog of Suspected Variable Stars, and derived new ephemerides for 2 stars already contained in the General Catalog of Variable Stars. Key words: Stars: variables: general – Surveys 1. Introduction Regular photographic observations of the sky for variable-star studies started in Moscow in 1895. Since then, several different telescopes were used to take direct sky plates for astrometry and for astrophysics. The Moscow plate archive now contains more than 60000 direct photographs and objective-prism plates taken in Moscow, at other sites in Russia, and at the Sternberg Institute’s observatory in Crimea, Ukraine. 280 A. A. The most important part of the Moscow plate collection are direct sky pho- tographs acquired in 1948–1996 with a 40-cm astrograph. This instrument was ordered by Prof. C. Hoffmeister for Sonneberg Observatory (Germany) and first installed there in 1938. 1658 plates from this telescope, taken in 1938–1945, are kept in Sonneberg (the GA series of the Sonneberg plate collection). In 1945, the telescope was taken to the Soviet Union as a part of the World War II reparations. It was initially installed in Simeiz (Crimea), then brought to Kuchino near Moscow, and in 1958 became the first instrument of the Crimean Laboratory of the Sternberg Institute in Nauchny, Crimea. The total number of plates taken with the 40-cm as- trograph after 1948 is about 22500. A single attempt of direct comparison between Sonneberg and Crimean plates of the 40-cm astrograph at a blink comparator was undertaken in 1980s (Samus 1983). The field of view of the 40-cm astrograph is 10◦ × 10◦ , on 30 × 30 cm plates (the focal length is 1600 mm). The typical exposure time for the variable-star fields was 45 minutes. The limiting magnitude of good-quality plates is about 17.5 (B). The instrument was mainly used for variable-star studies, including search for new variables. For some fields, rich series of plates exist (up to ≈ 500 plates). For variable stars that can be found in several fields, sometimes as many as 1000 pho- tographic plates are available. The list of fields, with numbers of plates obtained, can be found in Internet (http://cataclysm.sai.msu.ru/www/plates/40.dat). Plates are kept in good conditions, most plates, initially of excellent quality, are still per- fect. The Moscow plate collection, like other major astronomical plate collections of the world, has been actively used for scientific research for decades. It still contains a large amount of significant information never used by researchers, as indicated by discoveries of interesting events missed at the time of observation, like the discovery of Nova Aql 1985 (V1680 Aql) made 17 years later (Antipin et al. 2002). Guaranteed conservation of the vast amounts of information contained in the plate collection and its use by means of modern methods of image processing re- quire digitization of plate archives. This work commenced in Moscow, in 2004, after the purchase of two Creo EverSmart Supreme II scanners. The initial digiti- zation plans, along with a more detailed description of the Moscow plate archive from different instruments, were presented in Samus et al. (2006). Most plates from the 40-cm astrograph were taken for variable-star studies. It was natural to search for new variable stars using digital images obtained in the process of scanning the Moscow collection plates. In our first experiments, we discovered 38 new variable objects (mostly variable stars, but also extragalactic objects) on test partial scans (several square degrees) of star fields photographed with the astrograph (Sokolovsky 2006, Manannikov et al. 2006, Kolesnikova et al. 2007a,b). We introduced preliminary designations for variable stars discovered in this program with the prefix MDV (Moscow Digital Variable). Vol. 58 281 There were several other attempts to search for variable objects on digitized photographic plates. Among them are: a search for QSOs on the base of optical variability and zero proper motion criteria (Scholz et al. 1997, Brunzendorf and Meusinger 2001), a search for long-term variability using Sonneberg archival patrol plates (Vogt et al. 2004), a search for novae in M31 using Tautenburg Schmidt plates (Henze et al. 2008). In this paper, we announce the discovery and study of 274 new MDVs in the northern half of the field 66 Oph of the 40-cm astrograph. 2. Scanning and Reductions The field 66 Oph (18h 00.m 3, +4◦ 22′ , J2000.0) was photographed with the 40- cm astrograph in 1976–1995, a total of 254 plates were acquired. All these plates were scanned with a resolution of 2540 dpi (1.′′ 2 per pixel), pro- viding 14 bit per pixel per color. Color images produced by the scanner were saved in the TIFF (RGB) format using the scanner software operating in the Mac OS X environment. In our further reductions, we made use only of the green channel of each image, selected empirically. The files were then moved to a Linux server equipped with a 5 TB RAID array for storage and subsequent analysis. The images were converted to the FITS format using custom-written software∗. In this paper, we present our analysis of the northern half of the field (10◦ ×5◦ ) containing about 140000 stars within our detection limits (see below). The response to a point source of a given brightness on a large-scale photo- graphic plate is a subject to strong spatial variations. Obvious reasons for that include abberations in the optics of the astrograph (coma, vignetting, etc.), inho- mogeneity in photographic emulsion coating, and differences in airmass for stars in different parts of a plate. All these factors are expected to be relatively weak functions of coordinates on a plate. To overcome these complexities, the 10◦ × 5◦ field was subdivided into 72 nearly-square subfields. The influence of systematic factors is assumed to be the same for all stars in a given subfield. Each subfield was analyzed separately using VAST† software (Sokolovsky and Lebedev 2005), the results were combined at the final stage. For star detection and aperture photometry, VAST uses the well-known SEX- TRACTOR code (Bertin and Arnouts 1996). All objects identified by SEXTRAC- TOR as blended or non-point sources were excluded from further consideration be- cause such sources produce many false detections in a variability search. Aperture photometry was performed with a circular aperture. The aperture diameter was automatically selected for each image to compensate for seeing variations. This method was preferred against the variable elliptical aperture photometry (parameter MAG_AUTO) enabled by default in SEXTRACTOR, because the addition of extra ∗ftp://scan.sai.msu.ru/pub/software/tiff2fits †http://saistud.sai.msu.ru/vast 282 A. A. degrees of freedom (the aperture shape and size determined for each star separately) deteriorates the quality of measurements of faint stars. THE SEXTRACTOR param- eters and the aperture diameter were selected to optimize measurements of stars in the 13.5–16.5 mag in (B) range. This magnitude range was preferred because brighter variable stars in this particular field have mostly been already discovered in the ASAS-3 (Pojmanski´ 2002) and ROTSE-I/NSVS (Wo´zniak et al. 2004) CCD surveys, both covering the near-equatorial field of our plates. The VAST code automatically matches stars detected on an image by SEX- TRACTOR with stars detected on the reference image using the technique of the search for similar triangles. One of the best photographs was chosen as a reference image. Magnitudes of stars were measured by SEXTRACTOR in an instrumental scale with respect of the background level of the current image. All measured mag- nitudes were converted to the instrumental system of the reference image by ap- proximating the relation between magnitudes on the current and reference images with a parabolic function. All stars matched on the images were used to establish this relation. Visual inspection confirms that this approximation works well in the required range of magnitudes. The resulting light curves are characterized by an rms error of 0.05–0.15 mag for stars in the 13.5–16.5 mag range. 3. The Method of the Search for Variability and Its Limitations A light curve of a variable star is, obviously, characterized by a larger scatter of magnitude measurements compared to non-variable stars measured on the same series of images.